Bryce5.com

Registered users

» Forgot password
» Registration

Random image

the last entry
Comments: 3
tonkin

23.04.2024, 06:19

Home Page

Google ads below

Visible Median-cut

Visible Median-cut
Description:	This is inspired by David's test renders (Compair and Contrast). There are two ?Klein-Bottles? from David I've downloaded from the 3D Models section here, I cut them up. The scene is lit by the BlueFireCubeB-400, an HDRI I made from two artworks by - and for - David. It is a low contrast one like the MoonBlueGrey and works very nicely to show that the Specularity channel works under special circumstances with IBL light. The sphere at right has reflection set to 2.7 and shows the HDRI used and rendered as background. The other objects have specularity set to 100 (all other controls at 0), some also metallicity at 100. The white cut Klein-Bottle on the left has the default Bryce grey but all controls set to 0 except diffusion at 0.15. The cut Klein-Bottles on the ground have 0.07 diffusion in order to light up the otherwise completely black insides. The Backdrop would be completely white if it were not for the sphere I put in the camera's way. It has specularity at max, too, and shows how the light is mixed. The sphere in the centre (specularity and metallicity at 100) shows the lights generated by the median-cut algorithm. Rendered with 64 lights, intensity and effect both at 80.
Added by:	Horo
Keywords:	specularity, IBL, HDRI, median-cut, Klein-Bottle, davidbrinnen
Date:	05.02.2008 21:06
Hits:	3881
Downloads:	92
Rating:	0.00 (0 Vote(s))
File size:	223.3 KB

Previous image: Mugs

Next image: Compair and Contrast

Author:

Comment:

davidbrinnen
Admin

Join Date: 01.03.2004
Comments: 2224

Interesting. There is something curious about the way the specularity mapps to the surface of the geometries, some unexpectedly hard edges in there, in ways it is very similar to the diffusion, but not quite. Difficult to asses at this early stage how useful this effect might be, but what we have found so far, it shows some promise.

I'd be intrigued to learn how you cut up the bottle, when I did it I used wings3D and filled in the surface (to hide the double skin issue), here the incorrect double skin is quite visible. To be right, I need to figure out a way to make the bottle with a single surface instead of two. It is theoretically possible, put maybe beyond my megre wings moddling skills.

05.03.2008 18:21

Horo
Admin

Join Date: 05.26.2004
Comments: 4721

If your wings modelling skils are meagre, mine are picoscopic. I just booleaned half of the bottle off with a cube. There remain indeed a couple of questions about how the med-cut alg is really implemented; i.e. why amber and blue/cyan are mixed to green in a single one pixel light point. There's just no green in the HDRI.

Oh, well, on second thought, at least this is correct behavioure in the med-cut - it just occured to me. We not only learn, we also remember.

05.03.2008 19:16

rashadcarter1
Admin

Join Date: 06.04.2006
Comments: 2610

Hmm, the plot thickens. This is a very cool approach to solving for iron and perhaps other metals as the lack of diffusion is obvious to me. These are black bodies similar to cast iron. Objects rendered in this way appear to be very hard and solid. The brightness of the lights is obvious as I read the description and you say the diffusion is mere decimals for the white looking sphere. Wow, this effect requires very bright light.

In another way it makes me think of anisotropics. I have been testing ways to produce convincing velvet for a diamond scene and I have had to resort 100% to the methods explored above. I have been using image mapped reflection and bump with careful specualrity settings to control the reflection. To create the magical appearance of velvet requires a very careful setting of reflection, specularity, and bump. What is so different in the way of velvet as opposed to other shiny surfaces is that velvet is not wet and should not look so even though it shines under light. This is where a fabric image as bump information is essential.

Moreover, I have found a curious relationship between the specualrity channel and the reflection optic in a general sense. I have found that in some cases the reflection is actually controlled by the specualrity channels even when no specuarity is introduced. The experiement is very surprising to me.If you place an image into the reflection channel in node A(photo for sure but likely a procedural would have the same effect), you will notice that the reflection begins to look a bit like specularity. In working on the remakes of Ramon I have found that by using an image map as reflection it basically creates the effect of normal specularity. But even more interesting is that when you map reflection in this way the specular controls actually begin to affect the reflection. This is important to me because if you leave the specualr controls black your image mapped reflection may not look as you planned.

I do not understand the median-cut algorithm but I am guessing that it has the purpose of approximating colors. Clearly it is too much to expect that colors would be exact but the result is very pleasing above.

05.03.2008 23:55

Previous image:
Mugs

Next image:
Compair and Contrast

[Discord Server]